HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

[samsdad]

I have another thought experiment for you kids: let’s say that, in an alternate universe, the S&P500 behaved like UPRO behaves in this universe—we would think that the 16-20% CAGR of the “alternate universe S&P500” and its larger volatility were the norm; the benchmark; how the “market” just naturally behaved.

Would you still invest in the S&P 500 in this alternate universe?

This is flawed too. There are reasons the S&P 500 can't go to zero, or if it does we have much bigger problems. UPRO by design can go to zero even if the economy is fine longterm.

I am fine at 100/0 because I know with my timeline, if I really don't come out ahead some terrible things will be bigger worries.

I’m not sure that’s true, actually. We’ve already established that the circuit breakers/limit down mechanisms will stop all trading at -20% in a day. So it won’t do it like that. Over the course of a few days, I don’t think it will either, but it’d be close enough for our purposes. Someone mentioned it upthread. I’ll be back.

[MotoTrojan]

I have another thought experiment for you kids: let’s say that, in an alternate universe, the S&P500 behaved like UPRO behaves in this universe—we would think that the 16-20% CAGR of the “alternate universe S&P500” and its larger volatility were the norm; the benchmark; how the “market” just naturally behaved.

Would you still invest in the S&P 500 in this alternate universe?

This is flawed too. There are reasons the S&P 500 can't go to zero, or if it does we have much bigger problems. UPRO by design can go to zero even if the economy is fine longterm.

I am fine at 100/0 because I know with my timeline, if I really don't come out ahead some terrible things will be bigger worries.

I’m not sure that’s true, actually. We’ve already established that the circuit breakers/limit down mechanisms will stop all trading at -20% in a day. So it won’t do it like that. Over the course of a few dats, I don’t think it will either, but it’d be close enough for our purposes. Someone mentioned it upthread. I’ll be back.

Okay, it could go to well under 5%. Just look at 100% UPROSIM, I know you have the data. Or 100% the 2x fund that actually existed since the 90's. I don't believe that is even an argument, these equity funds will absolutely dig a hole that cannot be climbed out of if there is a significant crash, and thus will underperform the S&P500 long-term. It is the uncorrelation and rebalancing with LTT's that makes it so efficient.

[privatefarmer]

I have another thought experiment for you kids: let’s say that, in an alternate universe, the S&P500 behaved like UPRO behaves in this universe—we would think that the 16-20% CAGR of the “alternate universe S&P500” and its larger volatility were the norm; the benchmark; how the “market” just naturally behaved.

Would you still invest in the S&P 500 in this alternate universe?

This is flawed too. There are reasons the S&P 500 can't go to zero, or if it does we have much bigger problems. UPRO by design can go to zero even if the economy is fine longterm.

I am fine at 100/0 because I know with my timeline, if I really don't come out ahead some terrible things will be bigger worries.

It can only go to zero if it lost 33% in a day. We could all get nuked by North Korea, too. Both unlikely. SSO did go down by >80% but recovered in 4 years.

[MotoTrojan]

I have another thought experiment for you kids: let’s say that, in an alternate universe, the S&P500 behaved like UPRO behaves in this universe—we would think that the 16-20% CAGR of the “alternate universe S&P500” and its larger volatility were the norm; the benchmark; how the “market” just naturally behaved.

Would you still invest in the S&P 500 in this alternate universe?

This is flawed too. There are reasons the S&P 500 can't go to zero, or if it does we have much bigger problems. UPRO by design can go to zero even if the economy is fine longterm.

I am fine at 100/0 because I know with my timeline, if I really don't come out ahead some terrible things will be bigger worries.

It can only go to zero if it lost 33% in a day. We could all get nuked by North Korea, too. Both unlikely. SSO did go down by >80% but recovered in 4 years.

Fair enough I may have over-dramatized things. UPRO simulated data back to 1986 would've essentially tied the S&P500 but in doing so would've had a 97.5% drawdown in the process. I don't think any economy could survive a 97.5% reduction in expected or realized economic output and not implode.

[samsdad]

I have another thought experiment for you kids: let’s say that, in an alternate universe, the S&P500 behaved like UPRO behaves in this universe—we would think that the 16-20% CAGR of the “alternate universe S&P500” and its larger volatility were the norm; the benchmark; how the “market” just naturally behaved.

Would you still invest in the S&P 500 in this alternate universe?

This is flawed too. There are reasons the S&P 500 can't go to zero, or if it does we have much bigger problems. UPRO by design can go to zero even if the economy is fine longterm.

I am fine at 100/0 because I know with my timeline, if I really don't come out ahead some terrible things will be bigger worries.

I’m not sure that’s true, actually. We’ve already established that the circuit breakers/limit down mechanisms will stop all trading at -20% in a day. So it won’t do it like that. Over the course of a few dats, I don’t think it will either, but it’d be close enough for our purposes. Someone mentioned it upthread. I’ll be back.

Okay, it could go to well under 5%. Just look at 100% UPROSIM, I know you have the data. Or 100% the 2x fund that actually existed since the 90's. I don't believe that is even an argument, these equity funds will absolutely dig a hole that cannot be climbed out of if there is a significant crash, and thus will underperform the S&P500 long-term. It is the uncorrelation and rebalancing with LTT's that makes it so efficient.

Here's two quotes from vineviz, who's never been wrong if I recall correctly. Let's start with the upside, since that's why we're here:

I am not doubting or being negative, I am just want to know what conditions would need to be present for this strategy outperform the market?

It basically only takes two conditions for a daily leveraged strategy like this to outperform the market.

1) The underlying assets need to produce a modestly positive average return (i.e roughly 0.28% per month or better).
2) The underlying assets need to have slightly more up days than down days on average (i.e. at least one more up day per month than down days).

It's definitely true that a string of unfavorable months can definitely take a bite out this strategy: going back to 1970 a daily 3x leverage strategy like this would have produced three drawdowns greater than 50% (1972-74, 1977-1981, and 2007-2009) versus just one for an unleveraged stock portfolio.

Even so, this kind of leverage doesn't magnify the downside risk nearly as much as people tend to assume. The 3x 40/60 portfolio proposed here. Since 1970, the 20 worst months for large cap U.S. stocks have had an average return of -10.35%. During those same 20 worst months, this strategy would have had an average return of -13.47%.

On the downside:

* Can the leveraged ETFs be wiped out in a market crash? SSO (2x leveraged S&P) was in existence during 2008 and only dropped 81% while unleveraged S&P dropped 50%. I would have thought SSO would have been wiped out. Or does the leveraging only asymtotically approach zero?

Because the leverage on these ETFs resets daily they don't get wiped out due to market movements. If you took the five worst days for the S&P 500 from 1923 to present and experienced them in an uninterrupted streak a 3x daily leveraged ETF would lose only 91% of its value.

Needless to say, the odds of the S&P 500 dropping by 50% in five days is insanely low: I think the worst week in stock market history is a drop of less than 20%.

So, I'm not really too worried about it going to zero--or now that you've modified your position--"well under 5%" and never returning.



NOW, it might not return in my lifetime, but I've said from the start, this might be for my daughters and not me. And, we haven't gotten to the other half of the pie yet.

[MotoTrojan]

NOW, it might not return in my lifetime, but I've said from the start, this might be for my daughters and not me. And, we haven't gotten to the other half of the pie yet.

If you want to go 100% UPRO, go for it. I am fairly certain neither of us are saying that is as good idea so this is entirely a moot discussion .

[samsdad]

NOW, it might not return in my lifetime, but I've said from the start, this might be for my daughters and not me. And, we haven't gotten to the other half of the pie yet.

If you want to go 100% UPRO, go for it. I am fairly certain neither of us are saying that is as good idea so this is entirely a moot discussion .

Agreed. I’d never go 100% UPRO. If TMF flopped over dead tomorrow I’d hit up UBT. If that died too I’d go with TLT and reallocate accordingly.

[hdas]

Perhaps the fellow travelers in this thread can benefit from reading this paper:

Diversification Return and Leveraged Portfolios
Edward Qian
The Journal of Portfolio Management Summer 2012

Abstract

It is widely accepted that portfolio rebalancing adds diversification return to fixed-weight portfolios, but this is only true for long-only unleveraged portfolios. Qian provides analytical results regarding portfolio rebalancing and the associated diversification returns for different kinds of portfolios including long-only, long-short, and leveraged. He shows that portfolio rebalancing is linked to underlying portfolio dynamics. For long-only unleveraged portfolios, rebalancing amounts to a mean-reverting strategy, and the diversification return is always non-negative. But for short (or inverse) and leveraged portfolios, portfolio rebalancing on the top-down level amounts to a trend-following strategy that detracts from diversification return. Qian analyzes diversification returns of risk parity portfolios and shows that the diversification return of a leveraged long-only portfolio can generally be decomposed into two parts, both of which are related to a scaled unleveraged portfolio. The first part is the positive diversification return from rebalancing among individual assets at the bottom- up level, which is amplified by leverage. The second part is the negative diversification return caused by the leverage of the overall portfolio. His numerical examples show that diversification return is, in general, positive for leveraged risk parity portfolios when the leverage ratio is not too high. In addition, he shows that low correlations between different assets are crucial in achieving positive diversification return and reducing portfolio turnover for risk parity portfolios.

He has a very nice book as well, perhaps the best I read in the subject. Cheers

[MotoTrojan]

NOW, it might not return in my lifetime, but I've said from the start, this might be for my daughters and not me. And, we haven't gotten to the other half of the pie yet.

If you want to go 100% UPRO, go for it. I am fairly certain neither of us are saying that is as good idea so this is entirely a moot discussion .

Agreed. I’d never go 100% UPRO. If TMF flopped over dead tomorrow I’d hit up UBT. If that died too I’d go with TLT and reallocate accordingly.

Don’t forget EDV! Hopefully we won’t need them.

[MotoTrojan]

Perhaps the fellow travelers in this thread can benefit from reading this paper:

Diversification Return and Leveraged Portfolios
Edward Qian
The Journal of Portfolio Management Summer 2012

Abstract

It is widely accepted that portfolio rebalancing adds diversification return to fixed-weight portfolios, but this is only true for long-only unleveraged portfolios. Qian provides analytical results regarding portfolio rebalancing and the associated diversification returns for different kinds of portfolios including long-only, long-short, and leveraged. He shows that portfolio rebalancing is linked to underlying portfolio dynamics. For long-only unleveraged portfolios, rebalancing amounts to a mean-reverting strategy, and the diversification return is always non-negative. But for short (or inverse) and leveraged portfolios, portfolio rebalancing on the top-down level amounts to a trend-following strategy that detracts from diversification return. Qian analyzes diversification returns of risk parity portfolios and shows that the diversification return of a leveraged long-only portfolio can generally be decomposed into two parts, both of which are related to a scaled unleveraged portfolio. The first part is the positive diversification return from rebalancing among individual assets at the bottom- up level, which is amplified by leverage. The second part is the negative diversification return caused by the leverage of the overall portfolio. His numerical examples show that diversification return is, in general, positive for leveraged risk parity portfolios when the leverage ratio is not too high. In addition, he shows that low correlations between different assets are crucial in achieving positive diversification return and reducing portfolio turnover for risk parity portfolios.

He has a very nice book as well, perhaps the best I read in the subject. Cheers

Interesting. Through google I stumbled upon his Risk Parity book and almost bought it. Do you find this to be a better start (if you've read both)?

[privatefarmer]

Perhaps the fellow travelers in this thread can benefit from reading this paper:

Diversification Return and Leveraged Portfolios
Edward Qian
The Journal of Portfolio Management Summer 2012

Abstract

It is widely accepted that portfolio rebalancing adds diversification return to fixed-weight portfolios, but this is only true for long-only unleveraged portfolios. Qian provides analytical results regarding portfolio rebalancing and the associated diversification returns for different kinds of portfolios including long-only, long-short, and leveraged. He shows that portfolio rebalancing is linked to underlying portfolio dynamics. For long-only unleveraged portfolios, rebalancing amounts to a mean-reverting strategy, and the diversification return is always non-negative. But for short (or inverse) and leveraged portfolios, portfolio rebalancing on the top-down level amounts to a trend-following strategy that detracts from diversification return. Qian analyzes diversification returns of risk parity portfolios and shows that the diversification return of a leveraged long-only portfolio can generally be decomposed into two parts, both of which are related to a scaled unleveraged portfolio. The first part is the positive diversification return from rebalancing among individual assets at the bottom- up level, which is amplified by leverage. The second part is the negative diversification return caused by the leverage of the overall portfolio. His numerical examples show that diversification return is, in general, positive for leveraged risk parity portfolios when the leverage ratio is not too high. In addition, he shows that low correlations between different assets are crucial in achieving positive diversification return and reducing portfolio turnover for risk parity portfolios.

He has a very nice book as well, perhaps the best I read in the subject. Cheers

Interesting. Through google I stumbled upon his Risk Parity book and almost bought it. Do you find this to be a better start (if you've read both)?

has anyone found the book at a decent price?? $90 is a little steep...

[hdas]

Interesting. Through google I stumbled upon his Risk Parity book and almost bought it. Do you find this to be a better start (if you've read both)?

I didn't like his Risk Parity book. For RP I got Roncalli's, recommended to me by various ppl, but I haven't read it yet....

[MotoTrojan]

Interesting. Through google I stumbled upon his Risk Parity book and almost bought it. Do you find this to be a better start (if you've read both)?

I didn't like his Risk Parity book. For RP I got Roncalli's, recommended to me by various ppl, but I haven't read it yet....

Sweet. If you say this other one is the way to go I’ll think about it. Certainly steeper price.

[MotoTrojan]

HEDGEFUNDIE, any thoughts on the 1% correction factor for UPRO and 0.5% correction factor for TMF that siamond has found improved correlation?

[privatefarmer]

Meb Faber just basically talked about this strategy In his most recent podcast on the all weather portfolio. Not the exact same strategy but in the same ballpark. Even mentions at the end that you could use leverage to boost your returns although he doesn’t mention leveraged ETFs. Just a coincidence I think or maybe he’s following along in this thread

[Sanjuan]

I’m pondering doing this in my tax-deferred IRA at Fidelity (don’t have a Roth). Are these ETFs readily available? Is it as simple as selling current assets and buying UPRO/TMF? Any other hoops to go through?

[aj76er]

I’m pondering doing this in my tax-deferred IRA at Fidelity (don’t have a Roth). Are these ETFs readily available? Is it as simple as selling current assets and buying UPRO/TMF? Any other hoops to go through?

At Fidelity, I had to sign a "Designated Investments Agreement" to indicate that I was an experienced investor. This all occurred during the "Buy" order. Signing and reading over the agreement took about 5mins (the whole document is short). They also wanted me to update the profile for my Roth account, and I picked "Very Aggressive" (although I'm not sure it matters).

[HEDGEFUNDIE]

After weeks of work, Siamond has produced simulated 3X funds going back to 1955. Here is his data file.

First, to level set.

Here is the new simulated UPRO compared to the actual UPRO:





A 2% discrepancy over 10 years is not bad!

Here is the new simulated TMF compared to the actual TMF. The discrepancy in the early years (which also affected my original simulated TMF) remains; by 2011 the simulated TMF tracks the real TMF pretty well. If I had to guess, I would guess in the early years the spreads on the swaps were less advantageous to the fund than they are now.





Finally, here is a comparison of the strategy with the old simulated funds used in the Original Post (Portfolio 1) vs the new simulated funds (Portfolio 2) from 1987-2018.



The new, better simulated funds actually do slightly better than my original simulated funds.

[HEDGEFUNDIE]

Next, let's look at the new UPRO data compared to the S&P 500 going all the way back to 1955:



Of course, this ride would not have been for the faint-of-heart. Here is a comparison of the drawdowns:

[siamond]

Here is the new simulated TMF compared to the actual TMF. The discrepancy in the early years (which also affected my original simulated TMF) remains; by 2011 the simulated TMF tracks the real TMF pretty well. If I had to guess, I would guess in the early years the spreads on the swaps were less advantageous to the fund than they are now.

You're fast!

If you check the corresponding Telltale chart, you'll see that TMF had a hard time in the first couple of years. Several other leveraged funds had a similar 'eventful' start. There could be many reasons for that (size, experience, etc), but I do not think it is worth pondering too much about it. What matters is the trajectory of the fund vs. model once the fund found a more solid footing.

[HEDGEFUNDIE]

And now, for the main event, what will likely spark discussion for the next thousand posts, the performance of Long Term Treasuries.

Here is the performance of the new simulated TMF compared to unleveraged LTT:



Clearly, something happened around 1982 that fundamentally changed the dynamics of LTTs. (More on this later)

Here are the drawdowns:

[Kevin M]

Next, let's look at the new UPRO data compared to the S&P 500 going all the way back to 1955:
<snip>

I see you quickly modified your post to delete an erroneous statement about UPRO producing 6X the index return, along with an apparently erroneous telltale chart. Clearly not the case, and nothing to do with the logarithmic scale--just look at the final balance numbers.

Kevin

[HEDGEFUNDIE]

Now, what does all this mean for the strategy?

Here is the performance of the strategy since 1955 as compared to unleveraged S&P 500:



The strategy ends up with 2x the return of the unleveraged index, but of course that is not the whole story.

Here is the same comparison from 1955-1981:



And the same comparison from 1982-Present:

[HEDGEFUNDIE]

Next, let's look at the new UPRO data compared to the S&P 500 going all the way back to 1955:
<snip>

I see you quickly modified your post to delete an erroneous statement about UPRO producing 6X the index return, along with an apparently erroneous telltale chart. Clearly not the case, and nothing to do with the logarithmic scale--just look at the final balance numbers.

Kevin

Yeah, clearly a bug with PV.

[HEDGEFUNDIE]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term. Among the many academic articles written to support this view is this one:

https://www.nyu.edu/econ/user/gertlerm/qje00.pdf

Our estimates point to a significant difference in the way monetary policy was conducted pre- and post-late 1979. In the pre-Volcker years the Fed typically raised nominal rates by less than any increase in expected inflation, thus letting real short- term rates decline as anticipated inflation rose. On the other hand, during the Volcker-Greenspan era the Fed raised real as well as nominal short-term interest rates in response to higher expected inflation. Thus, our results lend quantitative support to the view that the anti-inflationary stance of the Fed has been stronger in the past two decades.

Now that all the data is in, who's in and who's out? And why?

[retiringwhen]

TLDR; leveraged Long-Term Treasuries is a slow poisonous death in a long-term secular rising interest environment.

[Kevin M]

Clearly, something happened around 1982 that fundamentally changed the dynamics of LTTs. (More on this later)

Yes--coupon rates started out at a historical high in 1982, and the general yield decreases since then added an average positive capital return component. Coupon rates were relatively low in 1955, and the generally increasing yields added an average negative capital return component to the total return.

For a simple model of rolling 10 year to 9 year annually (assuming flat yield curve between 9 and 10 year maturities), the arithmetic average income return for 1955-1981 was +5.8%, and the average capital return component was -2.5%. For 1982-2018 these numbers were +5.8% and +2.0% (interesting that the income return component was about the same).

It seems to me that this would explain at least part of it.

Kevin

[GrowthSeeker]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

One always must guard against a post hoc ergo propter hoc fallacy.
I can't think of any other explanation for the radically different performance in the two time periods, but that doesn't mean there isn't some other explanation.

[software]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

I would hope the discrepancy is fairly obvious. The transition from a rising interest rate environment to a falling interest rate environment and the 30 year bull market for long term treasuries began.

To me, that is the single biggest risk with this portfolio. The great results are based on a flawed assumption that LTT will return anything similar to what they did since 1982.

If I had a guess, considering where we are starting, the results will be far similar to the first half of your dataset.

[vineviz]

TLDR; leveraged Long-Term Treasuries is a slow poisonous death in a long-term secular rising interest environment.

That conclusion isn't consistent with the facts in evidence.

[vineviz]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

One always must guard against a post hoc ergo propter hoc fallacy.
I can't think of any other explanation for the radically different performance in the two time periods, but that doesn't mean there isn't some other explanation.

There might be another explanation, but the difference is definitely NOT due to differences in inflation or changes in interest rates. If there is another explanation besides Fed policy, I've never seen any suggestions that stand up to testing.

[HEDGEFUNDIE]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

I would hope the discrepancy is fairly obvious. The transition from a rising interest rate environment to a falling interest rate environment and the 30 year bull market for long term treasuries began.

To me, that is the single biggest risk with this portfolio. The great results are based on a flawed assumption that LTT will return anything similar to what they did since 1982.

If I had a guess, considering where we are starting, the results will be far similar to the first half of your dataset.

The question you need to ask is why interest rates stopped rising in 1982, and have been on a downward course ever since.

What I am saying is that the causal factor for the strategy’s outperformance is still in place, i.e. a generation of Central Bank economists trained on what the proper levers are to control inflation.

[MotoTrojan]

TLDR; leveraged Long-Term Treasuries is a slow poisonous death in a long-term secular rising interest environment.

That conclusion isn't consistent with the facts in evidence.

Care to clarify?

[MotoTrojan]

So who is going to do a 3x bear long treasury portfolio ? Or the original 40/60 portfolio but with 60% 3x bear LTT...

Certainly curious to hear peoples thoughts. I can see both sides of the coin. Doesn't seem unfathomable that we stay in a 1-6% range of interest rates over the next few decades with real bond returns in the 0-3% range. Not sure what forces would cause us to drive up to the extremes of the 70-80's.

[HEDGEFUNDIE]

So who is going to do a 3x bear long treasury portfolio ? Or the original 40/60 portfolio but with 60% 3x bear LTT...

That’s a great way to lose all your money during the next stock crash.

[MotoTrojan]

So who is going to do a 3x bear long treasury portfolio ? Or the original 40/60 portfolio but with 60% 3x bear LTT...

That’s a great way to lose all your money during the next stock crash.

Sticking to 100% LTT 3x bear from 55-83 seems like it would've been great at-least. These results give me healthy caution but not pulling out and still fascinated to learn/hear more from others and my own research.

[hdas]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

I would hope the discrepancy is fairly obvious. The transition from a rising interest rate environment to a falling interest rate environment and the 30 year bull market for long term treasuries began.

To me, that is the single biggest risk with this portfolio. The great results are based on a flawed assumption that LTT will return anything similar to what they did since 1982.

If I had a guess, considering where we are starting, the results will be far similar to the first half of your dataset.

The question you need to ask is why interest rates stopped rising in 1982, and have been on a downward course ever since.

What I am saying is that the causal factor for the strategy’s outperformance is still in place, i.e. a generation of Central Bank economists trained on what the proper levers are to control inflation.

With the new data 55-19, run a MC simulation (using block of years bootstrap...3-7 years).....let us see the output.

[hdas]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

The new data doesn't change my opinion, I would just point out that since the beautiful crash protection of Long Bonds is at its core a behavioral anomaly, I wouldn't be surprised if at least once in the next 30 years it doesn't hold up....and perhaps that's all you need to derail the excellent adventure. Btw, the main reason I don't do something like this, is because I believe there are better ways (but require more work), so I'm rooting for the fellow travelers that are venturing on this.

[retiringwhen]

TLDR; leveraged Long-Term Treasuries is a slow poisonous death in a long-term secular rising interest environment.

That conclusion isn't consistent with the facts in evidence.

Care to clarify?

My rebuttal to viveniz = 1955 to 1981 = long-term secular rising interest rate environment (4 to 14% in a nearly straight line)



approximately 97% total drawdown between 1955 and 1981



QED

[interestediniras]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

The new data doesn't change my opinion, I would just point out that since the beautiful crash protection of Long Bonds is at its core a behavioral anomaly, I wouldn't be surprised if at least once in the next 30 years it doesn't hold up....and perhaps that's all you need to derail the excellent adventure. Btw, the main reason I don't do something like this, is because I believe there are better ways (but require more work), so I'm rooting for the fellow travelers that are venturing on this.

What are the better ways, if you don't mind sharing?

[MotoTrojan]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

The new data doesn't change my opinion, I would just point out that since the beautiful crash protection of Long Bonds is at its core a behavioral anomaly, I wouldn't be surprised if at least once in the next 30 years it doesn't hold up....and perhaps that's all you need to derail the excellent adventure. Btw, the main reason I don't do something like this, is because I believe there are better ways (but require more work), so I'm rooting for the fellow travelers that are venturing on this.

What are the better ways, if you don't mind sharing?

I'd guess they are referring to using futures to eliminate the volatility decay of daily rebalancing.

[samsdad]

So with this new data, I think the bottom line for the strategy is this:

If you believe we will ever return to the Federal Reserve interest rate regime of pre-1982, then you should probably not pursue this.

If, however, you believe that the Fed post-Volcker is a fundamentally different place, reliant on market forces to influence interest rates, then this strategy is a great way to get rich over the long term.

Now that all the data is in, who's in and who's out? And why?

One always must guard against a post hoc ergo propter hoc fallacy.
I can't think of any other explanation for the radically different performance in the two time periods, but that doesn't mean there isn't some other explanation.

There might be another explanation, but the difference is definitely NOT due to differences in inflation or changes in interest rates. If there is another explanation besides Fed policy, I've never seen any suggestions that stand up to testing.

vineviz,

In another thread, stlutz said this:

I would suggest that the breakdown of the Bretton Woods system in the early 70s was more fundamental than the Paul Volcker appointment. That's not to understate the importance of Volcker in the early 1980s, but the way the Fed approaches policy now vs. then is very, very different.

Do you think that "the breakdown of the Bretton Woods system in the early 70s" could have at least contributed to "the radically different performance in the two time periods"?

[badapu]

Would addition /combination of intermediate term bonds (TYD) or REIT (DRN) have helped ride out the 1980s interest rate changes with less draw down? Also what about the option of using unleveraged long term tax exempt bonds VWLUX with UPRO?

[MotoTrojan]

Would addition /combination of intermediate term bonds (TYD) or REIT (DRN) have helped ride out the 1980s interest rate changes with less draw down? Also what about the option of using unleveraged long term tax exempt bonds VWLUX with UPRO?

I considered this, but the correlation would potentially be higher to equities and give less bonus there. Hedge suggested this but if you do want some exposure to leveraged 10 year (TYD) you should consider EDV instead; it is a Vanguard fund investing in 20-30 year strips which gives similar exposure without leverage costs/decay.

[HEDGEFUNDIE]

Would addition /combination of intermediate term bonds (TYD) or REIT (DRN) have helped ride out the 1980s interest rate changes with less draw down? Also what about the option of using unleveraged long term tax exempt bonds VWLUX with UPRO?

Siamond has posted the data for 2x and 3x S&P, ITT and LTT in the Simba spreadsheet:

https://drive.google.com/open?id=16ORud ... WfTP-0FggA

Feel free to generate some charts and post here if you find anything interesting.

[badapu]

I suspect we will have to go back to original "All Weather Portfolio" and tinker
30% in U.S. stocks
40% in Long-term U.S. Treasury Bonds
15% in Intermediate-Term U.S. Treasury Bonds
7.50% in Gold
7.50% in broad Commodity basket
This was modeled by Dalio to work well through the 1980 environment
Question is how much leverage was being applied to each piece

Commodities has negative corelation with long term treasury and US stock from 1972 to 2016
https://engineeredportfolio.com/2017/06 ... -or-reits/

I believe we are onto something and just needs a few more tweaks

[SVT]

I suspect we will have to go back to original "All Weather Portfolio" and tinker
30% in U.S. stocks
40% in Long-term U.S. Treasury Bonds
15% in Intermediate-Term U.S. Treasury Bonds
7.50% in Gold
7.50% in broad Commodity basket
This was modeled by Dalio to work well through the 1980 environment
Question is how much leverage was being applied to each piece

Commodities has negative corelation with long term treasury and US stock from 1972 to 2016
https://engineeredportfolio.com/2017/06 ... -or-reits/

I believe we are onto something and just needs a few more tweaks

I'd personally rather keep it simple (with less pieces), especially when I'm only throwing 10% to this strategy.

[gw]

To reiterate, from about 1000 posts ago....

Doesn't this strategy rely on a bond bull market?
No....

This is essentially BS. Whether you know it or not, the main reason you like this strategy is that long bonds have been an awesome investment over the last 30 years, which covers your backtesting horizon.

...

TLDR - if you think the appeal of this strategy isn't driven by the outstanding performance of bonds over the past 30 years, which is highly unreliable going forward, then you're simply fooling yourself.

And the rest of y'all need to stop reinvesting your retirement portfolios into 3x levered ETFs based on some anonymous internet dude's half-baked, week-old forum post. Jesus.

You can rationalize, but this thread would have died on page 1 if the original backtests included the 1955-81 period.

[StrangePenguin]

First post in this thread but I've been watching it with interest.

Seems like we've now got results for 2 contrasting eras: First is the long, slow runup in long-term interest rates from the 50s until early 80s, and the results aren't pretty. Then in the subsequent period from the early 80s till now, with a long, slow decline in long-term interest rates, the results are spectacular.

So the question I have is -- what if both of those periods are an anomaly? What if the future is that long-term interest rates remain approximately flat at, let's say 3-5% for the next few decades? Seems like the returns of this strategy would be somewhere in between the two eras discussed above (with the huge caveat that who knows that the stock market will do...).

[gtwhitegold]

After reviewing this, I am developing a stronger belief that it will probably be better to roll your own futures than to use leveraged ETFs. That or possibly reduce the leverage ratio and include other assets such as TIPS.